shapiro comment on draft vmgs · 2020-03-03 · shapiro comment on draft vertical merger guidelines...
TRANSCRIPT
Shapiro Comment on Draft Vertical Merger Guidelines
Page 1
Comment on DOJ/FTC Draft Vertical Merger Guidelines
Carl Shapiro†
3 March 2020
The Draft Vertical Merger Guidelines (“Draft VMGs”) issued by the DOJ and the FTC (“Agencies”) are
a significant improvement over the treatment of vertical mergers found in the 1984 Merger Guidelines.
The Draft VMGs are superior to the 1984 Merger Guidelines both (a) in their articulation of the
economic issues arising in vertical merger analysis, and (b) in their accuracy in describing how the
Agencies have actually analyzed vertical mergers for many years.
The Draft VMGs helpfully reference the Horizontal Merger Guidelines on a number of common issues.
This promotes consistency and predictability in merger enforcement.
In 2018, I testified as an economic expert on behalf of the DOJ in its unsuccessful challenge to the
vertical merger between AT&T and Time Warner.1 My comments here are based in part on that
experience. That experience also informed my presentation to the OECD at their June 2019 Roundtable
on Vertical Mergers in the Technology, Media, and Telecom Sector. I am appending to this comment
my submission to the OECD, “Testing Vertical Mergers for Input Foreclosure.”2 Although that
submission only addressed one theory of harm – input foreclosure – I hope that the step-by-step analysis
provided in my OECD submission will be helpful to the Agencies. See especially Figures 3 and 4.
I now provide brief additional comments on two parts of the Draft VMGs.
Market Definition and Market Shares
The Draft VMGs give prominence to the market shares of the merging parties.
Market shares are clearly informative in horizontal merger analysis, where they form the basis of the
structural presumption. However, their role in vertical merger analysis is entirely different. The Draft
VMGs do not properly or fully reflect this basic economic observation.
Section 3 of the Draft VMGs states:
“The Agencies are unlikely to challenge a vertical merger where the parties to the merger have a
share in the relevant market of less than 20 percent, and the related product is used in less than
20 percent of the relevant market.”
The Draft VMGs do not provide an economic basis for these market share thresholds, which appear to
be arbitrary. The Agencies should explain their bases for these 20 percent thresholds. Does this approach
reflect recent practice or does it represent a change in practice? The public deserves to know.
I am concerned that using market shares in this manner could well weaken vertical merger enforcement.
† Professor at the University of California at Berkeley. See http://faculty.haas.berkeley.edu/shapiro.
1 My Expert Report is available at https://www.justice.gov/atr/case-document/file/1081336/download. My Rebuttal Report is
available at https://www.justice.gov/atr/case-document/file/1081321/download.
2 My OECD submission is directly available at https://one.oecd.org/document/DAF/COMP/WD(2019)75/en/pdf.
Shapiro Comment on Draft Vertical Merger Guidelines
Page 2
Share of the Relevant Market
Suppose that Firm A sells an input that is used by Firm B, and these two firms propose to merge. In this
setting, Section 2 of the Draft VMGs defines a relevant market and identifies “related products.”
Consider the common case in which the relevant market is the downstream market in which Firm B
participates and the “related product” is the input supplied by Firm A.
In the AT&T/Time Warner case, Time Warner played the role of Firm A and AT&T played the role of
Firm B. The relevant product market was the distribution of multichannel video programming to
consumers. One “related product” was the Turner Content owned by Time Warner.
The fundamental problem with the way that the Draft VMGs use market shares is that they simply
assume that harm to downstream customers is less likely if the merged firm controls a smaller share of
the relevant downstream market. For those accustomed to analyzing horizontal mergers, this assumption
might seem natural and reasonable, but market shares play an entirely different role in vertical merger
analysis than they do in horizontal merger analysis. In the AT&T/Time Warner case, applying the same
methods described in Section 5 of the Draft VMGs to assess unilateral effects, greater harm to
consumers was predicted in local areas where AT&T’s downstream market share was smaller. This
resulted because the impact of AT&T raising its rivals’ costs was greater in those local areas.3
Share of the Relevant Market that Uses the Related Product
The Draft VMGs state: “The Agencies are unlikely to challenge a vertical merger where … the related
product is used in less than 20 percent of the relevant market.” This statement requires clarification.
First, the Agencies should clarify that evaluating the likely economic effects of a proposed vertical
merger does not require defining a relevant market containing the “related product.” Doing so often will
be unnecessary and could well be misleading. In the AT&T/Time Warner case, a key issue was whether
AT&T’s downstream rivals in the distribution of video programming to consumers would be
significantly harmed if they lost access to the Turner Content. There was direct evidence on that point,
making it unnecessary to define an upstream market containing the Turner Content.4 Doing so would
have been an uninformative distraction. The VMGs should state this explicitly.
Third, if the Agencies choose to make inferences based on whether “the related product is used in less
than 20 percent of the relevant market,” they should be much clearer and more explicit about how this
concept will be measured. The draft VMGs refer to the “share of the relevant market that uses the
related product.” The VMGs should explicitly state that this is not a measure of market share, either in
the relevant market or in another market containing the “related product.” The VMGs should explicitly
state that adding up this measure across different inputs used in the relevant market (which the merging
firms will presumably point to as substitutes) typically results in a total far in excess of 100 percent.
Again, the AT&T/Time Warner case is instructive. One can ask what percent of the (downstream)
relevant market used the related product, the Turner Content. The evidence showed that more than 90
3 Section 6 of my Expert Report, “Impact on Consumers in Local Areas,” states: “All else equal, subscribers in Zones with
more rival MVPDs are likely to observe higher post-merger prices because of the greater impact of the increase in rivals’
costs.” See p. 68 and Figure 18.
4 See Section 6 of my Rebuttal Report, “The Commercial Significance of the Turner Content.” AT&T’s economic expert,
Professor Dennis Carlton, questioned the importance of the Turner Content based on his measure of Turner’s share of
“television video content consumption.” I stated (p. 8): “In my view, much if not all of what Professor Carlton has to say
about what he calls the “video content market” is either uninformative or misleading for the purposes of evaluating how the
merger will affect the bargaining between Turner and MVPDs that compete against DTV.”
Shapiro Comment on Draft Vertical Merger Guidelines
Page 3
percent of multichannel video subscribers had access to the Turner Content.5 This reflected the
importance of Turner Content (including CNN) to consumers, and hence to multichannel video
distributors. The Turner Content was used in over 90 percent of the relevant market. I am concerned that
many people who are accustomed to thinking in term of market shares will be confused by the language
in the draft VMGs, because they will expect such measures to add up to 100 percent. In the AT&T/Time
Warner case, several other firms also supplied content that was used by a large share of subscribers,
including Disney, Fox, NBCUniversal, and Viacom. Adding up the share of the relevant market used by
each of these content providers, one would obtain a figure far in excess of 100 percent, perhaps 400 or
500 percent. To avoid confusion, the VMGs should explicitly make this point and provide an example.
Elimination of Double Marginalization (“EDM”)
Section 6 of the Draft VMGs states: “The agencies generally rely on the parties to identify and
demonstrate whether and how the merger eliminates double marginalization.” I support an approach
under which EDM, like all other efficiencies, must be shown to be cognizable before it can be credited.
My OECD submission explains how to place EDM within the overall analysis in cases involving input
foreclosure. Here, I emphasize three points relating to the treatment of EDM.
• As with other efficiencies, EDM must be shown to be merger specific to be credited. As stated in
my OECD submission (p. 6): “That is a factual question that must be assessed on a case-by-case
basis. For example, if other firms in the industry have managed to eliminate double
marginalization through contract, perhaps by using two-part tariffs or other non-linear pricing
schemes, the merging firms might well be able to do likewise. In that case, EDM is not merger
specific and should not be credited as an efficiency in the merger analysis.”
• Some aspects of the analysis of EDM are different from the analysis of other claimed merger
efficiencies. As stated in my OECD submission (p. 6): “EDM is treated differently from other
claimed merger synergies because a vertical merger inherently gives the merged firm an
incentive to set the downstream price based on the merged firm’s combined upstream and
downstream profits. Put differently, EDM follows logically from the normal working assumption
of antitrust economists that for-profit firms are run to maximize their overall profits. This is
directly analogous to our normal working assumption that a horizontal merger eliminates
competition between the merging parties and typically creates some upward pricing pressure.”
• When quantifying EDM, it is critical to include the opportunity cost to the merged firm when its
downstream division gains sales at the expense of its rivals. This opportunity cost arises because
the upstream division loses margins on its sales to those rivals. Not accounting for this
opportunity cost can lead one to greatly overstate the magnitude of EDM, especially if the input
sold by the upstream division is widely used by downstream rivals. Paragraph 24 of my OECD
submission provides a numerical example based on the AT&T/Time Warner merger.
5 See Section 3.2.1 of my Expert Report.
Organisation for Economic Co-operation and Development
DAF/COMP/WD(2019)75
Unclassified English - Or. English
3 June 2019
DIRECTORATE FOR FINANCIAL AND ENTERPRISE AFFAIRS
COMPETITION COMMITTEE
Testing Vertical Mergers for Input Foreclosure - Note by Carl Shapiro
Roundtable on Vertical mergers in the technology, media and telecom sector
7 June 2019
This paper by Carl Shapiro was submitted as background material for Item 10 of the 131st OECD
Competition committee meeting on 5-7 June 2019.
The opinions expressed and arguments employed herein do not necessarily reflect the official
views of the Organisation or of the governments of its member countries.
More documents related to this discussion can be found at
http://www.oecd.org/daf/competition/vertical-mergers-in-the-technology-media-and-
telecomsector.htm
JT03448343
This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the
delimitation of international frontiers and boundaries and to the name of any territory, city or area.
2 │ DAF/COMP/WD(2019)75
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
Testing Vertical Mergers for Input Foreclosure*
Professor Carl Shapiro
1. Investigating mergers is like auditing tax returns. We expect (hope?) that most
taxpayers honestly pay their taxes, but some do not. Detecting deficient tax returns is
important, to fairly collect the taxes that are owed and to deter tax evasion. Likewise, most
mergers do not pose a threat to competition, but some would cause substantial harm to
competition if allowed to proceed.1 Identifying those mergers is important, both to protect
the competition process and to deter anti-competitive mergers. This is true of vertical
mergers, even though as a category they threaten competition less than do horizontal
mergers.
2. This submission explains a practical procedure for analyzing vertical mergers to
identify those that are most likely to lessen competition and harm customers based on
“input foreclosure.” The input foreclosure theory of harm is identified and described in the
Background Note by the Secretariat, “Vertical Mergers in the Technology, Media, and
Telecom Sector.”2 The distinct “customer foreclosure” theory of harm is economically
analogous and can be developed and applied in suitable cases a parallel fashion to the one
outlined here. This submission does not address theories of horizontal collusion or loss of
potential competition that can arise in vertical mergers.3
3. The analysis here should prove useful even for a competition authority that
considers almost all vertical mergers to be pro-competitive, because it is applicable to
whichever vertical mergers the competition authority believes raise sufficient concerns
about input foreclosure to warrant some investigation. The analysis is deliberately
* Prepared for the OECD Competition Committee session “Vertical Mergers in the Technology,
Media, and Telecom Sectors, 7 June 2019. This is a preliminary draft subject to revision. Please
send comments to [email protected]. Carl Shapiro is a Professor at the University of California
at Berkeley. He served as the chief economist at the Antitrust Division of the U.S. Department of
Justice during 1995-1996 and 2011-2013, and as a Member of the President’s Council of Economic
Advisers during 2011-2012. He also has served as an economic consultant for the government and
for private parties on a variety of antitrust matters over the years. His papers and his curriculum vitae
are available at his web site, http://faculty.haas.berkeley.edu/shapiro.
1 According to the Fiscal Year 2017 FTC/DOJ Hart-Scott-Rodino Annual Report, 2052 transactions
were reported that year, of which 51 (2.6%) received a Second Request for additional information
and 39 (1.9%) involved an enforcement action.
2 See https://one.oecd.org/document/DAF/COMP(2019)5/en/pdf, especially in Section 4.1
“Foreclosure (unilateral effects).”
3 If the merging firms may become direct rivals in the foreseeable future, their merger is (also) a
horizontal merger involving the loss of potential competition. For a recent discussion of how to
handle mergers that may eliminate a potential competitor to a dominant firm, see Carl Shapiro,
“Protecting Competition in the American Economy: Merger Control, Tech Titans, Labor Markets,”
at http://faculty.haas.berkeley.edu/shapiro/protectingcompetition.pdf. For an analysis of potential
competition cases focused on innovation, see Giulio Federico, Fiona Scott Morton, and Carl Shapiro,
“Antitrust and Innovation: Welcoming and Protecting Disruption,” available at
http://faculty.haas.berkeley.edu/shapiro/disruption.pdf.
DAF/COMP/WD(2019)75 │ 3
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
structured to provide “off-ramps,” so a vertical merger investigation can quickly be closed
if the evidence indicates that harm to competition is unlikely.
4. I use the recent merger between AT&T and Time Warner to illustrate how one can
perform this type of analysis.4 I testified on behalf of the Department of Justice in that
merger.5 AT&T, primarily through its DirecTV service, was a major distributor of video
content to households. Time Warner owned a valuable collection of video content, the
“Turner Content.” AT&T distributed the Turner Content to Pay TV Households, as did
AT&T’s major competitors, which are referred to as “multichannel video program
distributors” or MVPDs. The leading MVPDs were the cable companies Comcast and
Charter and the Dish direct broadcast satellite service. The input foreclosure theory of harm
involved the merged entity weakening its downstream MVPD rivals by raising the price of
Turner Content.6
5. Vertical mergers differ fundamentally from horizontal mergers in that they do
involve a loss of direct competition between the merging the parties. In the United States,
as a practical matter, it is much harder for the DOJ and the FTC to challenge a vertical
merger in court than a horizontal one, because there is no simple route by which they can
establish a presumption that a vertical merger is likely to harm competition. In horizontal
mergers, the structural presumption plays that role.
6. Vertical mergers are inherently more complex to analyze than horizontal mergers
because one must consider two levels of competition. The step-by-step analysis described
below is based on well-established economic principles, and the elements of this analysis
have been used in previous cases. I believe the framework offered here is as simple as
possible while providing reliable and useful results. Nonetheless, the judge in the
AT&T/Time Warner case mocked the analysis as overly complex
1. Elements of the Analysis of Input Foreclosure
7. The analysis in this submission seeks to determine whether a proposed vertical
merger is likely to lessen downstream competition due to input foreclosure and thus harm
final consumers. The mechanism of harm to competition is that the merged entity would
use its control over the acquired upstream input to weaken its downstream rivals, either by
denying them access to that input – “total foreclosure” – or by raising the price charged for
4 The AT&T/Time Warner merger is very briefly described in the Background Note (p.17).
5 My detailed expert report is available at https://www.justice.gov/atr/case-
document/file/1081336/download. My rebuttal expert report is available at
https://www.justice.gov/atr/case-document/file/1081321/download. The District Court sharply
criticized my analysis; United States v AT&T, Case 1:17-cv-02511 (DDC, 2017). On appeal, the
District Court decision was found not to be clearly erroneous; United States v. AT&T, Case 18-
5214, (DC Circuit, 2019). My intention here is not to re-litigate that case, but rather to use it to
illustrate how the input foreclosure theory of harm can be developed in practice. Rather obviously,
the AT&T/Time Warner case also illustrates some of the difficulties and risks facing a competition
authority seeking to challenge a vertical merger.
6 In the language of the Background Note (¶41), the case involved partial foreclosure. Economic
analysis indicated that it would not be profitable for the merged entity to engage in total foreclosure
by refusing to license the Turner Content to rivals.
4 │ DAF/COMP/WD(2019)75
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
that input – “partial foreclosure.” For simplicity, I will refer to both of these effects in
terms of “raising rivals’ costs” (RRC).
8. Figure 1. displays the elements of the analysis of vertical mergers in which the
competition concern involves the input foreclosure theory of harm. (All figures are at the
end of the submission.). The term “EDM” refers to “elimination of double
marginalization,” a common efficiency from vertical mergers.
9. Figure 2. shows these elements as they presented themselves in the AT&T/Time
Warner merger.
2. Step-by-Step Analysis: Ability and Incentive
10. The rest of this submission describes a step-by-step analysis to identify and analyze
the input foreclosure theory as applied to a proposed vertical merger. The step-by-step
analysis outlined here fits comfortably within the three-step ability-incentive-effect
framework described in Section 3 of the Background Note.
11. Figure 3. provides a flow chart for the “ability “and “incentive” parts of the
analysis.
12. The first step is to ask whether the upstream input involved in the merger is
competitively significant to the downstream rivals that have been using that input.7 An
input is competitively significant to a downstream rival if that firm’s ability to compete
would be substantially impaired if were to lose access to the input. In this step, one should
consider these rivals’ best response to that lack of access, which could involve using a
substitute input or modifying their products or services. These responses will generally
leave the rivals with higher costs and/or lower quality products or services.
13. In the AT&T/Time Warner case, there was extensive evidence about the
importance of the Turner Content for MVPDs. For starters, every major MVPD distributed
Turner Content to almost all of its subscribers, and Turner had been able to significantly
raise the price it charged for its content over time, notwithstanding the movement of some
customers away from Pay TV packages and toward streaming. Turner Content was
generally seen as one of the most important packages of content for MVPDs to carry. There
was some hotly disputed evidence specifically about the impact on an MVPD of losing
access to the Turner Content. However, there was no such evidence from an actual, long-
term blackout of Turner Content on an MVPD. In the end, the District Court was not
convinced by the estimates I used about the share of subscribers an MVPD would lose due
to the absence of the Turner Content.
14. The next step is to ask whether the merged entity’s downstream operations would
gain a substantial volume of business if one or more downstream rivals lost customers
because they were denied access to the input in question or raised their price due to higher
7 The Background Note describes a similar concept at ¶25, asking if the merging parties control as
asset that is important to rivals and unique, meaning it is difficult to replace with alternatives.
DAF/COMP/WD(2019)75 │ 5
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
input costs.8 The key metric in this step is the diversion ratio from these downstream rivals
to the merged entity.9
15. In the AT&T/Time Warner case, I used local market shares to estimate the
diversion from an MVPD losing access to the Turner Content toward DirecTV. I also
accounted for the fact that some of these lost subscribers would not sign up with any
MVPD. In the end, the District Court was not convinced by the estimates I used about the
diversion ratio from rival MVPDs to DirecTV.
16. The third step is to ask whether the business diverted from the downstream rivals
to the downstream operations of the merged firm generates significant profit margins for
the merged entity. The key metric in this step is the pre-merger price/cost margin at the
downstream merging firm.
17. In the AT&T/Time Warner case, the downstream Pay TV packages sold by
DirecTV to consumers were generally large bundles that had substantial margins. The DOJ
had difficulty obtaining accurate and timely margin information from AT&T. I used the
best data available to me, which was a moving target during the litigation. In the end, the
District Court was not convinced by the margin estimates I used.
18. If quantitative analysis is possible, the calculations performed will vary from case
to case, depending on whether total foreclosure or partial foreclosure is being considered.
But these are relatively minor details. From an economic perspective total foreclosure is
just a special (and extreme) case of partial foreclosure. The quantification also will vary
depending on whether (a) the upstream suppliers set the input prices and downstream firms
respond, or (b) the input prices are negotiated. But again these are relatively minor details,
variations to fit and reflect the institutional facts of the case.
19. The core, robust economic principle is that a vertical merger raises the economic
cost to the merged firm of selling the input to its rivals, because access to the input, or lower
prices for that input, make them stronger competitors. This core idea is intuitive to business
executives. As customers, business executives often are wary and feel vulnerable when
they must buy an important input from a rival. Likewise, as integrated suppliers, business
executives commonly are tempted to use a key input strategically to advantage their
downstream operations. In many cases, business documents will reflect these basic
economic and strategic ideas. The analysis above seeks to assess, on a case-by-case basis,
whether these anti-competitive incentives, which are inherent to vertical mergers, are
substantial. This is not unlike using the upward pricing pressure (UPP) metric to assess, on
a case-by-case basis, whether the inherent anti-competitive incentives in a horizontal
merger are substantial.
8 The rival downstream firms also can experience cost increases because alternative upstream input
suppliers raise their prices in response to the input foreclosure by the merged firm.
9 The Background Note describes a similar concept at ¶29, with reference to the diversion ratio from
the rival to the merged firm and to the price/cost margin at price/cost margin of the merged firm.
Vertical merger analysis is similar to horizontal merger analysis in that both involve the
multiplicative product of a diversion ratio and a price/cost margin.
6 │ DAF/COMP/WD(2019)75
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
3. Step-by-Step Analysis: Efficiencies and Effects
20. In cases where a proposed vertical merger would create significant incentives for
input foreclosure, the analysis then proceeds to consider efficiencies that might offset these
incentives. Figure 4. provides a flow chart for this “effects” part of the analysis. If some
efficiencies are cognizable, meaning that they are merger-specific and verified, then some
balancing is needed to determine the merger’s effects.
21. The analysis of efficiencies begins by considering the elimination of double
marginalization (“EDM”). EDM is a well-known economic aspect of vertical mergers. One
good way to think about EDM is to recognize that, after the merger, starting from pre-
merger prices, the downstream division of the merged entity will have some incentive to
lower price, if that lower price will attract more customers and if those extra customers
generate extra profits at the upstream division of the merged firm.
22. EDM is treated differently from other claimed merger synergies because a vertical
merger inherently gives the merged firm an incentive to set the downstream price based on
the merged firm’s combined upstream and downstream profits. Put differently, EDM
follows logically from the normal working assumption of antitrust economists that for-
profit firms are run to maximize their overall profits. This is directly analogous to our
normal working assumption that a horizontal merger eliminates competition between the
merging parties and typically creates some upward pricing pressure. In econ-speak, the
difference is that a horizontal merger internalizes a negative pecuniary externality from
lower prices while a vertical merger internalizes a positive pecuniary externality from lower
prices.
23. However, it is important to recognize that, while we must assume that a vertical
merger will lead to the elimination of double marginalization, this does not imply that EDM
is merger-specific. That is a factual question that must be assessed on a case-by-case basis.
For example, if other firms in the industry have managed to eliminate double
marginalization through contract, perhaps by using two-part tariffs or other non-linear
pricing schemes, the merging firms might well be able to do likewise. In that case, EDM is
not merger specific and should not be credited as an efficiency in the merger analysis. In
the AT&T/Time Warner case, my analysis credited EDM as merger-specific.
24. In cases where EDM is determined to be merger-specific, one may be able to
estimate the magnitude of EDM. EDM can be measured as the gap between the pre-merger
price that the downstream merging firm is paying for the input and the economic marginal
cost to the upstream firm of providing that input to the downstream firm. Critically, this
economic marginal cost includes opportunity cost. The key question is how the upstream
division is affected when the downstream division lowers its price to attract more
customers. To see how this works, consider the polar case in which all of these extra
customers were previously purchasing from downstream rivals who were using the
upstream division’s input. In that case, the upstream division gains no new sales when the
downstream firm lowers its price, so the magnitude of the EDM effect is zero. More
generally, if the upstream division’s price/cost margin is M and if a fraction X of customers
attracted by the downstream firm were not already using the upstream input, then the EDM
is not M but only M times X. In the AT&T/Time Warner case, I calculated EDM as $1.20
per subscriber per month. Purely for illustrative purposes, this figure would result if the
price/cost margin on the Turner Content was $6 and if 20% of the new subscribers attracted
to DirecTV by a price decrease would be new Turner customers. It would be a major error
DAF/COMP/WD(2019)75 │ 7
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
to use the Turner price/cost margin of $6 as the magnitude of EDM; that would overstate
EDM by a factor of five.
25. If the RRC effect and the EDM effects can be quantified, they can be compared.
This amounts to asking whether the merger will raise or lower the costs of the input for the
downstream firms collectively. The merging downstream firm enjoys an input cost
reduction, due to EDM, but the rivals downstream firms suffer from higher input costs. In
the AT&T/Time Warner merger I determined that RRC was greater than EDM.10 This
conclusion was hotly disputed, and the District Court judge did not find my analysis reliable
enough to conclude that the merger would lead to higher costs and thus harm competition
and consumers. Since the burden of proof was on the DOJ, this proved fatal to the DOJ’s
case.
26. If RRC is greater than EDM, the proposed merger could still promote competition
and benefit customers if the merger would lead to significant cognizable synergies. As with
horizontal mergers, some healthy skepticism about claimed efficiencies is warranted, and
the burden rests on the merging parties to establish that they are merger specific and verified
and ultimately will lead to consumer benefits.
4. Conclusions and Lessons
27. This submission has sketched out a practical, step-by-step analysis for evaluating
vertical mergers to see whether they pose a threat to competition based on input foreclosure.
This approach could form an important part of updated Vertical Merger Guidelines, which
are badly needed in the United States.
28. The core economic idea behind this approach is intuitive: when one firm acquires
an important input, that firm will have an incentive to use that input to favor itself over its
rivals. Business executives understand this point well. Offsetting this, the merger may
allow for greater vertical coordination. The framework here provides a structural method
of assessing those two basic effects.
29. Nothing in this analysis hinges on market definition, market shares, or measures of
market concentration. In this respect, the analysis of vertical mergers is entirely unlike the
analysis of horizontal mergers, where these elements, and the structural presumption, can
be critical.11
30. For vertical mergers that generate significant merger-specific efficiencies, the best
outcome for competition and for consumers may be for the merger to proceed with suitable
remedies, which may include behavioral remedies. The efficacy of structural and
behavioral remedies should be assessed on a case-by-case basis. Ruling out behavioral
10 I also developed a downstream merger simulation model to estimate how these cost changes for
the Turner Content would be passed through to final consumers. I used that model to predict
consumer harm. That model also was disputed.
11 In fact, in the AT&T/Time Warner merger, I did not find greater harms in geographic markets
where DirecTV had a larger share of the MVPD market. To see why, consider the polar case in
which the downstream firm has 100% of the downstream market. In that case, if EDM is merger
specific, the merger will generate benefits for consumers and will not raise rivals’ costs, because
there are no rivals. Of course, such a merger may still harm competition by raising downstream entry
barriers.
8 │ DAF/COMP/WD(2019)75
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
remedies in vertical mergers a priori is unwarranted. Adopting such a rigid approach may
force a competition authority to accept one of two inferior outcomes: allowing the merger
to proceed without any remedy or trying to block the merger in its entirety.12
Figure 1.
12 For example, in the NBC/Universal merger, the DOJ and the FCC allowed the merger to proceed
subject to behavioral remedies. I was the chief economist at the DOJ when that merger was reviewed.
In the AT&T/Time Warner case, AT&T argued that the remedy in the NBC/Universal merger had
been effective. The DOJ was unable to rebut that argument.
DAF/COMP/WD(2019)75 │ 9
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
Figure 2.
Figure 3.
10 │ DAF/COMP/WD(2019)75
TESTING VERTICAL MERGERS FOR INPUT FORECLOSURE - NOTE BY CARL SHAPIRO Unclassified
Figure 4.